Key Takeaways
- Google Cloud's operating margin reached 32.9% in Q1 2026, up from 17.8% a year earlier, on revenue of $20,028M (+63%) and operating income of $6,598M.
- Alphabet's capex was $35,674M in Q1 2026 against $17,197M a year earlier. Full-year 2026 guidance was raised to $180-190 billion, roughly double FY2025's $91.4 billion.
- Capex ran at 32.5% of Alphabet's Q1 2026 revenue of $109,896M, while depreciation was only $6,482M.
- Management guided Cloud's margin lower, expecting a low-single-digit percentage-point headwind for the rest of 2026 from the Wiz acquisition.
The margin went up, not down
An earlier version of this story circulated under the headline that AI scale was hitting Alphabet's margins. The data said the opposite then, and says the opposite now with more force.
In Q1 2026 Google Cloud produced $20,028M of revenue and $6,598M of operating income. That is a 32.9% operating margin, against 17.8% in Q1 2025. CFO Anat Ashkenazi put it plainly on the call: "Cloud operating income was $6.6 billion, tripling year over year, and operating margin increased from 17.8% in the first quarter of last year to 32.9%." The chart above traces the same series through Q4 2024 (17.5%) and Q4 2025 (30.1%).
Revenue grew 63% and exceeded $20 billion for the first time. Operating income roughly tripled. Margin expanded by 15.1 percentage points. This happened in the same quarter Alphabet spent $35,674M on property and equipment — more than double the $17,197M of a year earlier.
Operating leverage, and an accounting lag that looks like operating leverage
Two things are happening at once, and separating them is the whole analytical task.
The first is genuine operating leverage. Cloud revenue grew 63% while the cost base that serves it grew more slowly, which is what a scaling infrastructure business is supposed to do. Utilisation of an expensive fixed asset rose, and margin followed.
The second is timing. Capital expenditure does not hit the income statement when the cash leaves. It arrives later, spread across the asset's useful life, as depreciation. In Q1 2026 Alphabet spent $35,674M and recognised $6,482M of depreciation on property and equipment. Depreciation is running at roughly 18% of current capex.
That gap is not fraud or manipulation — it is how accrual accounting works. But it means a portion of today's 32.9% margin is a function of the fact that the servers bought this quarter have barely begun to depreciate. As the fleet matures, depreciation converges toward the run-rate of capital spending, and the margin faces a headwind that has nothing to do with demand. Alphabet's annual depreciation was $21.1 billion in FY2025 against $15.3 billion in FY2024; capex over the same period went from $52.5 billion to $91.4 billion.
The capital intensity number
Alphabet spent 32.5% of its total Q1 2026 revenue ($109,896M) on property and equipment. For a company whose consolidated operating margin is 36.1%, capital spending now consumes almost as much revenue as operating profit generates.
Full-year guidance was raised on 29 April to "$180 to $190 billion, up from our previous estimate of $175 to $185 billion, to now include investment related to the acquisition of Intersect, which closed in March." Against FY2025 capex of $91.4 billion, the midpoint implies a doubling in a single year. Alphabet is not alone: Amazon has told shareholders it expects to invest about $200 billion in 2026, and Microsoft's CFO guided to roughly $190 billion for the calendar year — an industry-wide shift in the cost of serving software that shows up in gross margins as well as in fixed assets.
The backlog is real, and it is not what it was
Cloud backlog reached $462 billion at the end of Q1 2026, against total remaining performance obligations of $92.4 billion a year earlier. A fivefold increase is the kind of number that ends arguments.
It should not, because the definition changed. Ashkenazi attributed the jump to "strong demand for our enterprise AI offerings and the inclusion of TPU hardware sales." Backlog that includes hardware sale agreements is not comparable to a prior period's backlog of cloud service commitments. The growth is large and probably still large after adjustment — but the reported multiple is not a like-for-like measure, and treating it as one is exactly the error that the wider capex cycle invites.
What would falsify the operating-leverage reading
The thesis here is that cloud margin expansion reflects durable operating leverage rather than an accounting artefact. Three things would undercut it.
If depreciation growth outpaces cloud gross profit growth over the next four quarters, the margin gain was substantially a timing effect. Depreciation rose 44% year on year in Q1 2026, against 63% cloud revenue growth — the margin held because revenue is still outrunning the depreciation catch-up, and that race is the thing to measure.
If margin falls without a corresponding revenue slowdown, the incremental capital is not earning its cost. Management has already signalled the first leg of this: they expect a low-single-digit percentage-point headwind to Cloud's operating margin for the remainder of 2026 from the Wiz acquisition. A margin that peaks at 32.9% and settles near 30% would still be a transformed business.
The segment disclosure also cannot answer the question that matters most. Alphabet does not break out the return on the AI infrastructure specifically, and cloud operating income includes the whole segment. There is no way, from published figures, to isolate what the $180-190 billion is earning. The margin tells you the segment is profitable; it does not tell you the marginal dollar of capex is.
The line that decides this
The interesting series from here is not margin and not revenue. It is depreciation, which grew 44% year on year while capex grew 107%. Those two rates cannot diverge indefinitely. When depreciation growth catches capex growth — a mechanical certainty on a five-to-six-year asset life, absent perpetual acceleration in spending — several points of cloud margin come out of the income statement without a single customer leaving. That is the moment the operating-leverage story gets tested, and it is coming into view on a schedule set by an accounting policy rather than by demand.
This content is for informational purposes only and does not constitute financial advice. Always do your own research or consult a qualified financial adviser before making investment decisions.